Salmon and Protected Areas

Adelaide Robinson

Background and Motivation

• Many populations of anadromous salmon are expected to go extinct within the next 100 years
  • 5 out of seven steelhead DPS (Distinct Population Segments) in California are federally listed
  • Both ESU (evolutionary significant units) of coho are listed
• While numerous factors affect their decline habitat loss is a major factor
• Urban land use has been linked to the decline over time and shown to impact juvenile survival in the Pacific Northwest

Question

How do protected areas impact salmon populations within California?

Data

• California Monitoring Plan for Salmon and Steelhead

  • CDFW and NOAA

  • Monitor and Assemble data on anadromous salmon populations across CA

    • population counts for watersheds across California

    • Geospatial data containing the extents of monitored watersheds

• California Protected Area Database

  • polygons of areas protected for open space across Ca

Protected Area Visualization

Analysis

\(populationcount = B_0 + B_1year_t + + B_2percentprotected + B_3Year_t * percentProtected +E_i\)

• Used OLS Linear Regression

• Used linear model with multiple group fixed affects

  • helps account for difference between populations and inconsistent collection at sites over all years

  • sweeps away variation between populations

  • I also used clustered standard error-accounts for the fact that error within populations will be correlated (violating heteroskedacticity)

Results

term	estimate	std.error	statistic	p.value	conf.low	conf.high
year	33.0267267	11.9953359	2.753297	0.0073366	9.1458473	56.9076061
percent_protected	24.8850371	17.6041939	1.413586	0.1614622	-10.1622206	59.9322949
year:percent_protected	-0.5845462	0.2058474	-2.839706	0.0057560	-0.9943569	-0.1747355

• Interpretation
  • Percent protected has a statistically insignificant affect on the number of fish.
  • It appears that percent protected decreases the rate of change in populations over time
    • -0.5 is the difference in the effect of year on fish count for every one increase in protected area
  • Year has a positive affect on the number of steelhead
    • when %protected is 0 on average there will be an increase of 33 fish per year
• Limitations
  • Not convinced my data is linear in parameters
  • Is there omitted variable bias?
  • Using a loose definition of protected
• What’s Next?
  • Would be better to look at overwinter survival rates
  • Explore the relationship between protected areas and survival in a different way